Question After 20 years of building, finally a dead PSU

[Brainonska511]

It finally happened: a dead PSU. I was working in my home office which I share with my spouse, and I hear her computer click and shuts off. It wouldn't reboot (just had the power light stuck on) - we unplug it and plug it back in, boots fine and seems okay. But 2 hours later, it happens again, except this time, that trick doesn't work. I pulled the PSU (a 2-year-old Corsair RM750e) and hooked it into a tester - nothing. Fortunately, I had an RM550x hanging around from a prior iteration of my own computer to pop in as a spare. Now I just have to see if Corsair will come through with the RMA.

 

[BoomerD]

2 years old? Corsair should take care of that immediately. They've always been great for me when something dies.

 

[igor_kavinski]

I SHOULD have had a dead PSU when I was running an overclocked RX 580 on a 400W PSU with Core i7-3770. It was a Huntkey and recommended PSU for stock operation of that card is 450W. I think the PSU still works though haven't needed to use it in years.

 

[Brainonska511]

2 years old? Corsair should take care of that immediately. They've always been great for me when something dies.

We'll see. They responded to my ticket and have offered the ATX 3.1 version of the RM750e (because they don't have the 2023 version). Just waiting for an update after I affirmed that the replacement was okay.

 

[mindless1]

I had a wave of PSU fail that were bought between approx 2005 and 2010, lasting about 2-6 years each, but they were mediocre units, budget tier 600W and lower, mostly 420W to 500W, $20 or less after rebate, makes like Ultra, XFX, Coolermaster, and Corsair's low end at the time, CX430.

One of the CX430 held on until a couple years ago but wasn't in an everyday-use system. That was the only one I've had fail in the last decade.

Almost all of them had capacitors failing. I recapped a couple CX430 and maybe one of the XFX, and a few more that escape my memory at the moment, but many weren't worthy.

I'm only counting PSU I bought myself, whether for self-use or for someone's system I was building on the cheap or repairing, not PSU I didn't purchase that were in systems people brought to me before I learned to say NO! If I were to count those too, one of the other common failure points was Yate Loon sleeve bearing fans in Sparkle aka FSP Fortran PSU. Their caps also popped, and a few transistors too, but that was after lack of airflow baked them.

 

[mikeymikec]

For PCs I built for myself I think I've only had one PSU die, a no-name job that came with the case in my first PC build. IIRC a power surge nailed the PSU and board. BX chipset. I started using surge-protection mains extension strips after that. It was some time later that I started buying better (or at least branded) power supplies, and a bit more time after that before I started reading reviews for PSUs!

 

[igor_kavinski]

Their caps also popped, and a few transistors too, but that was after lack of airflow baked them.

I'm using a Seasonic fanless 600W Titanium PSU with my 9950X3D build. How long do you think it will last without direct airflow?

 

[mindless1]

I'm using a Seasonic fanless 600W Titanium PSU with my 9950X3D build. How long do you think it will last without direct airflow?

Impossible to make a prediction with any expectation of accuracy.

Measure PSU internal temperature while at average and peak (extended) load. The rest of the system's airflow is going to matter a lot, how much positive or negative case pressurization the case I/O fans cause to flow through it, as well as efficiency vs load, how much heat is produced.

Generally speaking, I avoid passive PSU because even a 120mm fan spinning at the lowest RPM it can do without stalling, which is usually inaudible when it's a sleeve bearing fan, is going to result in it running cooler than a passively cooled PSU in "most" cases.

Granted I am not considering some of the modern cases that strap on several 120mm fans every place they can, if most of those are intake fans (besides one in the rear, and depending on size and RPM) then the positive case pressurization is going to force a good amount of air out the PSU.

 

[igor_kavinski]

The rest of the system's airflow is going to matter a lot, how much positive or negative case pressurization the case I/O fans cause to flow through it, as well as efficiency vs load, how much heat is produced.

It's installed in the worst way possible. It's facing downward because I didn't want the heat from it flowing upwards into the case (yet some of it still will end up doing that). I only used it coz I didn't want to buy one more PSU. I have almost seven now already, including two really lame old ones that shouldn't really be used with modern systems.

 

[Brainonska511]

We'll see. They responded to my ticket and have offered the ATX 3.1 version of the RM750e (because they don't have the 2023 version). Just waiting for an update after I affirmed that the replacement was okay.

Corsair finally approved my RMA late Sunday night and provided a prepaid label. Just need to find a box now and collect all the modular cables (kind of annoying that I couldn't just keep those). They say a replacement will be sent after they've processed the return.

 

[BoomerD]

Corsair finally approved my RMA late Sunday night and provided a prepaid label. Just need to find a box now and collect all the modular cables (kind of annoying that I couldn't just keep those). They say a replacement will be sent after they've processed the return.

Toldja...

 

[mindless1]

It's installed in the worst way possible. It's facing downward because I didn't want the heat from it flowing upwards into the case (yet some of it still will end up doing that). I only used it coz I didn't want to buy one more PSU. I have almost seven now already, including two really lame old ones that shouldn't really be used with modern systems.

Flip it over and strap an intake fan on the outside?

I always set up cases to have positive pressurization for filtered intake, so if it were flipped over in that situation, the heat would still be pulled down and out instead of rising into the system.

 

[FelixDeCat]

Built my first PC around 1990/1, 486DX2. I had a few power supply fails.

For a long time, power supplies were trash, especially the "free" ones that came with cheap $39 cases from Frys.... you get what you pay for.

My only recent fail was also a brand new Corsair back around 2014. Replaced it with RMA then gave it away for a better supply. My current Corsair has been running perfectly since 2017....just before Frys became a ghost town.

If you remember, Frys actually technically shut down with no new stock and a few idle employees in darkened buildings with 'token' shelve stock TWO YEARS before the 2020 pandemic. Then they declared bankruptcy and liquidated.

I miss Frys.

 

[Fallen Kell]

Corsair finally approved my RMA late Sunday night and provided a prepaid label. Just need to find a box now and collect all the modular cables (kind of annoying that I couldn't just keep those). They say a replacement will be sent after they've processed the return.

They wanted the modular cables because the new one might come with new pin layout and/or new wire gauge requirements and they don't want you using the old cables that may not meet the standards, causing some other issue/ticket and possibly some headache with you pointing out that they didn't require the old cables be sent back and/or didn't specifically warn you that the old cables could cause problems, and thus they are somehow liable.

 

[Brainonska511]

They wanted the modular cables because the new one might come with new pin layout and/or new wire gauge requirements and they don't want you using the old cables that may not meet the standards, causing some other issue/ticket and possibly some headache with you pointing out that they didn't require the old cables be sent back and/or didn't specifically warn you that the old cables could cause problems, and thus they are somehow liable.

Perhaps. I believe both use type 4 cables though, so it shouldn't have been a problem. But I certainly understand that perspective.